campaigns to motivate families (3). If we are to overcome the "tragedy of the commons"* described by Hardin (4), then ways are needed to persuade individuals that their own interests and those of their families, as well as the interests of the community or society at large, require limitation of family size. Similarly, Taylor, in his suggestions for a "five stage population policy" (5), emphasizes the need to go beyond the population that is merely waiting for services and attract those who are not so fully convinced. In both of these cases, knowledge of the consequences of excessive numbers at the family level would be invaluable.
PRESENT KNOWLEDGE ABOUT THE EFFECTS OF FAMILY SIZE AND BIRTH INTERVAL
There is a substantial body of evidence concerning the relation between family size, or number of children, and birth interval and a number of factors relevant to health. This evidence was obtained by a diverse array of investigators studying a variety of problems over a period of many years. The studies were carried out in many countries, in all stages of economic development, and among various social classes within a given country. The approaches varied: some were retrospective (based on currently obtained data describing past events), others cross-sectional, still others prospective or longitudinal (identifying children at birth and following them to see what happens).
Each health indicator that has been examined in these studies is produced by complex, and usually obscure, interactions of numerous causal factors. There is undoubtedly a considerable amount of overlap—the same set of interacting causal factors involved in producing various effects. In no case is the total interaction clearly understood, but certainly family size and birth interval operate only as parts of the causal web. In the great majority of these studies, family size and birth interval were among many factors examined as independent variables in relation to a given problem, and not the primary concern of the investigators. Therefore, the interaction between family size, for example, and other relevant variables, such as socioeconomic status is often unexamined. In spite of this, the general pattern of effects suggested by the evidence available is so consistent that it seems reasonable to consider that deficiencies in some studies are, in a sense, compensated for by the adequacy of other studies. We need not, in other words, be compelled to disregard the findings of an investigator who did not control for a given
*To illustrate the fact that a given act may have consequences that seem beneficial to the individual but are harmful to society at large, Hardin uses as an example "the commons," or common pasture. A fanner with two cows adds one more; he gets 50 percent more milk, but his one cow may be enough to push the total population of cows to such a level that the commons may be permanently damaged by overgrazing. The parallel problem with regard to population is obvious.